基于散射计波浪参数反演模型的改进算法实验

发布时间：2019-06-08 17:20

【摘要】：一种新的利用散射计(ERS1/2)数据反演有效波高和平均周期的模式被提出。通过俄罗斯学者利用浮标数据建立完全成长风浪条件下有效波高与风速之间的关系,与匹配浮标观测的有效波高数据对比,区分完全成长风浪、成长风浪和涌浪3种海况下的匹配数据;利用BP神经网络建立模式反演3种海况下的有效波高,均方根误差分别为0.53、0.57和0.86m,反演平均周期均方根误差分别为0.69、1.04和1.36s。这种反演方法在完全成长风浪海况下最好,依次是成长风浪和涌浪海况。该研究为散射计数据反演波浪参数提供了依据,使大面积反演波浪参数成为可能。
[Abstract]:A new model for inversion of effective wave height and average period using scattering meter (ERS1/2) data is proposed. The relationship between effective wave height and wind speed under the condition of complete growth wind wave is established by Russian scholars using buoy data, and the relationship between effective wave height and wind speed under the condition of complete growth wind wave is established, and compared with the effective wave height data observed by matching buoy, the complete growth wind wave is distinguished. The matching data of growing wind wave and surging wave under three kinds of sea conditions; Using BP neural network model to retrieve the effective wave height under three sea conditions, the root mean square error is 0.53m, 0.57m and 0.86m, respectively, and the average periodic root mean square error is 0.69, 1.04 and 1.36s, respectively. This inversion method is the best under the condition of fully growing wind and sea, followed by growing wind and wave and surging sea condition. This study provides a basis for the inversion of wave parameters from scattering meter data, and makes it possible to invert wave parameters in large areas.
【作者单位】：中国科学院烟台海岸带研究所山东省海岸带环境过程重点实验室;南京信息工程大学海洋科学学院;Department
【基金】：中国科学院与乌俄白国际合作项目“理论数字模拟海况对散射计海面风场反演精度的影响”(Y429011031) 国家自然科学基金委员会与俄罗斯基础研究基金会合作交流项目“利用SAR数据分析关于雷达后向散射及风场反演的影响因子”(4141101049),国家自然科学基金委面上项目“多波段极化合成孔径雷达岸高风场反演研究”(41176160)
【分类号】：P715.7

【相似文献】